RESEARCH

Risk of post-pregnancy hypertension in women with a history of

hypertensive disorders of pregnancy: nationwide cohort study
Ida Behrens,1 Saima Basit,1 Mads Melbye,1 Jacob A Lykke,2 Jan Wohlfahrt,1 Henning Bundgaard,3
Baskaran Thilaganathan,4 Heather A Boyd1

1
Department of Epidemiology ABSTRACT pregnancy. Rates in women with a hypertensive
Research, Statens Serum OBJECTIVES disorder of pregnancy were threefold to 10-fold higher
Institut, Artillerivej 5, DK-2300 To determine how soon after delivery the risk of post- 1-10 years post partum and remained twice as high
Copenhagen S, Denmark
2 pregnancy hypertension increases in women with even 20 or more years later.
Department of Obstetrics,
Copenhagen University Hospital hypertensive disorders of pregnancy and how the risk CONCLUSIONS
(Rigshospitalet), Copenhagen, evolves over time. The risk of hypertension associated with hypertensive
Denmark
3 DESIGN disorders of pregnancy is high immediately after an
The Heart Centre, Copenhagen
University Hospital Nationwide register based cohort study. affected pregnancy and persists for more than 20
(Rigshospitalet), Copenhagen, SETTING years. Up to one third of women with a hypertensive
Denmark disorder of pregnancy may develop hypertension
4
Denmark.
Fetal Medicine Unit, within a decade of an affected pregnancy, indicating
StGeorges Hospital, University POPULATIONS
of London, London, UK
that cardiovascular disease prevention in these
482972 primiparous women with a first live birth
Correspondence to: H A Boyd
women should include blood pressure monitoring
or stillbirth between 1995 and 2012 (cumulative
hoy@ssi.dk initiated soon after pregnancy.
incidence analyses), and 1025118 women with at
Additional material is published
online only. To view please visit
least one live birth or stillbirth between 1978 and
the journal online. 2012 (Cox regression analyses).
Cite this as: BMJ 2017;358:j3078 MAIN OUTCOME MEASURES Introduction
http://dx.doi.org/10.1136/bmj.j3078
10 year cumulative incidences of post-pregnancy Hypertensive disorders of pregnancy (pre-eclampsia;
Accepted: 17 June 2017 hypertension requiring treatment with prescription eclampsia; haemolysis, elevated liver enzymes, and
drugs, and hazard ratios estimated using Cox low platelets (HELLP) syndrome; and gestational
regression. hypertension) affect up to 10% of pregnancies.1 2
Women with a hypertensive disorder of pregnancy
RESULTS
have increased risks of post-pregnancy hypertension,
Of women with a hypertensive disorder of pregnancy
in a first pregnancy in their 20s, 14% developed ischaemic heart disease, and stroke,3-5 which has
hypertension in the first decade post partum, prompted changes to guidelines by the American
compared with 4% of women with normotensive Heart Association and the European Society of
first pregnancies in their 20s. The corresponding Cardiology to include hypertensive disorders of
percentages for women with a first pregnancy in their pregnancy as risk factors for cardiovascular disease
40s were 32% and 11%, respectively. In the year in women.6 7 However, clinical awareness of the link
after delivery, women with a hypertensive disorder between hypertensive disorders of pregnancy and
of pregnancy had 12-fold to 25-fold higher rates of cardiovascular disease is incomplete,8-10 which may
hypertension than did women with a normotensive result in delayed diagnosis and jeopardise the health
of these women. Furthermore, it is unclear how soon
after an affected pregnancy screening for hypertension
WHAT IS ALREADY KNOWN ON THIS TOPIC and other markers of cardiovascular disease should be
 omen with a history of hypertensive disorders of pregnancy have a twofold to
W initiated.2 6 11 Recent work suggests that the immediate
fourfold increased post-pregnancy risk of developing essential hypertension, an postpartum years are important, with one study
important risk factor for cardiovascular disease reporting a fivefold increase in hypertension rates in
the first five years after a pre-eclamptic pregnancy,12
It is unclear how soon after delivery the risk of hypertension increases in women
and another finding that 25-45% of women with
with hypertensive disorders of pregnancy
a hypertensive disorder of pregnancy developed
It is also unclear how the risk changes with time since pregnancy, and therefore, hypertension within five years of delivery.13 However,
there is no evidence on which to base recommendations for clinical follow-up of precisely when the increased risk of hypertension
these women appears after a pregnancy affected by a hypertensive
WHAT THIS STUDY ADDS disorder of pregnancy and how the risk changes with
time since pregnancy remain unclear. Understanding
Women had substantially increased risks of post-pregnancy hypertension in
patterns of hypertension risk after a hypertensive
the decade after a hypertensive disorder of pregnancy (14-32% after a first
disorder of pregnancy would enable clinicians to plan
pregnancy, depending on age at delivery)
appropriate post-pregnancy follow-up and diagnose
Cardiovascular disease prevention in women with hypertensive disorders hypertension as early as possible.
of pregnancy should include blood pressure monitoring initiated soon after In a cohort of more than one million women delivering
pregnancy in Denmark in 1978-2012, we examined the timing and

thebmj|BMJ 2017;358:j3078 | doi: 10.1136/bmj.j3078 1

 RESEARCH
trajectory of post-pregnancy hypertension risk on a fine
temporal scale, in women with and without a history
of hypertensive disorders of pregnancy. Specifically,
we estimated cumulative incidences of post-pregnancy
hypertension over the first 10 years post partum and
compared rates of post-pregnancy hypertension in
women with and without a hypertensive disorder of
pregnancy, by time since most recent pregnancy.
Methods
Data sources
The Danish civil registration system continuously
updates personal and vital status information through
the unique personal identification number assigned
to all Danish residents.14 Contacts with the healthcare
system and filled prescriptions are registered using
the personal identification number, enabling register
based studies with little loss to follow-up. The
national patient register contains information on
hospital discharge diagnoses assigned since 1977
and outpatient diagnoses assigned since 1995.15 The
medical birth register contains information on live
births and stillbirths since 1973, with gestational age at
delivery from 1978.16 The national prescription register
contains data on prescriptions filled since 1994.17
Study cohorts
In this study we used two cohorts: one for the
estimation of cumulative incidence of post-pregnancy
hypertension and the other cohort for the estimation of
hazard ratios for post-pregnancy hypertension.
Cumulative incidences of post-pregnancy hypertension
using the medical birth register, we identified all women
with a first pregnancy lasting 20 or more weeks and
ending in live birth or stillbirth between 1995 and 2012.
We excluded women with any cardiac or circulatory
system disorder (international classification of diseases,
eighth revision (ICD-8), codes 390-458, or 10th revision
(ICD-10) codes I00.0-I99.9) registered in the national
patient register before their first delivery, and women
with known or potential pregestational hypertension.
Hazard ratios for post-pregnancy hypertension
using the medical birth register, we identified all
women who had at least one pregnancy that lasted
20 or more weeks and ended in live birth or stillbirth
between 1978 and 2012, and who were living in
Denmark at some point during the follow-up period,
1995-2012. We excluded women with any cardiac or
circulatory system disorder registered in the national
patient register before their first delivery or before
1995, whichever came later, and women with known
or potential pregestational hypertension.
Hypertensive disorders of pregnancy (exposure)
We considered women to have a hypertensive
disorder of pregnancy in a given pregnancy if they
had a diagnosis of gestational hypertension, pre-
eclampsia, eclampsia, or HELLP syndrome registered
in the national patient register any time between one
month before delivery and seven days post partum.
To be considered exposed to a hypertensive disorder
2
of pregnancy in our study, women with diagnoses
registered outside this time window also had to have at
least one diagnosis registered within the window. (We
adopted this restriction to try to ensure that diagnoses
of hypertensive disorders of pregnancy reflected true
cases; we judged that assigning a hypertensive disorder
of pregnancy designation would be questionable if
a womans only hypertensive disorder of pregnancy
diagnoses were registered outside this time window,
particularly if a diagnosis was assigned many weeks
before delivery and never alluded to again.) Since by
definition a hypertensive disorder of pregnancy involves
incident hypertension in a pregnant woman with onset
after 20 weeks gestation, we considered only diagnoses
registered after this point. As registered in the national
patient register, gestational hypertension is defined
as hypertension without accompanying proteinuria
(ICD-8 code 637.00, ICD-10 code O13.9 or O16.9),
whereas in moderate pre-eclampsia, mild or moderate
hypertension is accompanied by proteinuria (ICD-8
codes 637.03, 637.09, or 637.99, ICD-10 code O14.0
or O14.9). Severe pre-eclampsia fulfills the criteria for
moderate pre-eclampsia, with the addition of one or
more of severe hypertension, severe proteinuria, signs
of organ failure (including the HELLP syndrome), or
generalised seizures (ICD-8 codes 637.04, 637.19,
762.19, 762.29, or 762.39, ICD-10 codes O14.1,
O14.2, or O15.0-15.9). If a woman was registered with
more than one hypertensive disorder of pregnancy in a
single pregnancy, we classified her as having the most
severe disorder registered. In an additional analysis,
we used an alternative categorisation for hypertensive
disorders of pregnancy whereby we classified the
condition as early onset for deliveries at less than 34
weeks gestation, intermediate onset for deliveries at
34-36 weeks, and term for deliveries at 37 or more
weeks, regardless of the severity of the registered
hypertensive disorder of pregnancy.
Hypertension (outcome)
We considered a woman to have new onset post-
pregnancy hypertension from the time she filled
a second prescription for antihypertensive drugs
(Anatomic Therapeutic Chemical codes C02-03 or
C07-09 registered in the national prescription register)
within a six month period. When defining hypertension
we ignored antihypertensive drug use that was
potentially related to treatment for hypertensive
disorders of pregnancy (use from 20 weeks before
delivery to three months post partum).
Exclusion of women with known or potential
pregestational hypertension
In the cumulative incidence analyses we considered
women with antihypertensive drug use before their
first pregnancy or up to 20 weeks gestation in that first
pregnancy to have pregestational hypertension and
excluded them from the cohort.
In the relative risk analyses we excluded women
with pregnancies before 1994 (ie, predating the
national prescription register) who filled prescriptions
doi: 10.1136/bmj.j3078|BMJ 2017;358:j3078 | thebmj

for antihypertensive drugs in 1994, as we did not
know whether hypertension was pregestational or
post-gestational in women using drugs at the start
of follow-up. Similarly, we excluded women with a
first pregnancy in 1994 if they filled prescriptions for
antihypertensive drugs before 20 weeks gestation or
3-12 months after delivery (ie, beyond the period when
such treatment could be associated with pregnancy
related hypertension) since we did not know if
this use began before or after pregnancy. We also
excluded women with a first pregnancy after 1995 and
antihypertensive drug use before 20 weeks gestation
in that pregnancy.
Covariates
Age, maternal birth year, parity, diabetes, smoking,
and body mass index were considered a priori to be
potential confounders of an association between
hypertensive disorders of pregnancy and later
hypertension. Information on age and maternal birth
year (civil registration system), parity (medical birth
register), and diabetes (types 1 and 2, national patient
register, ICD-8 codes 249.00-249.09 or 250.00-
250.09, ICD-10 codes E10.0-E11.9) was available for
the entire study period. Information on first trimester
smoking status and prepregnancy body mass index
(medical birth register) was available from 1991 and
2004 onwards, respectively.
Statistical analyses
To calculate 10 year cumulative incidences of
hypertension, we followed women from three
months after their first delivery until the first of:
hypertension, 10 years after the first delivery, death,
emigration, missing in the civil registration system,
or 31 December 2012. For women with at least two
pregnancies, we also conducted similar analyses for
the decade after the second delivery.
For the estimation of relative risks of post-pregnancy
hypertension we followed women from 1 January
1995 or three months after their first delivery in the
study period, whichever came later, until the first of:
hypertension, death, emigration, missing in the civil
registration system, or 31 December 2012. Using Cox
regression with maternal age as the underlying time, we
estimated hazard ratios for post-pregnancy hypertension
by history of hypertensive disorders of pregnancy.
Firstly, we estimated hazard ratios by number of years
since a womans most recent pregnancy, to examine
the immediate effect of a hypertensive disorder of
pregnancy on subsequent risk of hypertension. In these
analyses, hypertensive disorder of pregnancy status
was a time dependent variable reflecting a womans
experience in the most recent pregnancy and was
updated for each subsequent pregnancy. Secondly, we
estimated hazard ratios by most severe hypertensive
disorder of pregnancy to date, stratified by attained
age. History of hypertensive disorders of pregnancy
was a cumulative time dependent variable, reflecting
the most severe hypertensive disorder of pregnancy (if
any) the woman had experienced to date. A womans
thebmj|BMJ 2017;358:j3078 | doi: 10.1136/bmj.j3078
RESEARCH
status could switch from, for example, no hypertensive
disorder of pregnancy to pre-eclampsia if the pre-
eclamptic pregnancy came later, but she could not
revert to having no history of a hypertensive disorder
of pregnancy if an unaffected pregnancy followed
an affected one. Finally, in women with at least two
pregnancies who were hypertension-free between
pregnancies, we estimated hazard ratios comparing
rates of hypertension in women with a hypertensive
disorder of pregnancy in the first pregnancy, in the
second pregnancy, or in both pregnancies, with rates in
women with two normotensive pregnancies, by number
of years since the second pregnancy. For women with
more than two pregnancies, follow-up in these analyses
ended at their third pregnancy.
In all Cox regression analyses, we stratified the
baseline hazards by parity (1, 2, 3 live births and/
or stillbirths) and maternal birth year (five year
intervals) to ensure that we compared rates in women
of the same age and parity; including maternal birth
year helped to account for possible time trends in
hypertensive disorders of pregnancy and hypertension
treatment. Analyses estimating hazard ratios by most
severe hypertensive disorder of pregnancy to date were
also adjusted for time since most recent pregnancy
(<1 year, 1 year intervals from 1-9 years, 10-14 years,
15-19 years, and 20 years). In sensitivity analyses in
the full cohort, we additionally adjusted for diabetes
as a time dependent variable (such that we considered
women without diabetes at baseline who developed
diabetes during follow-up to be without diabetes until
their diagnosis, after which they contributed person
time to the diabetes group). In subcohorts of women
with available information, we further adjusted for
smoking and prepregnancy body mass index. Smoking
status (smoker or non-smoker) and prepregnancy body
mass index (<18.5, 18.5-24, 25-29, 30-34, or 35)
were considered as time independent variables based
on information from the womans first pregnancy in
or after 1991 and 2004, respectively. We evaluated
violation of the proportional hazards assumption by
plotting Martingale residuals against attained age.18
All analyses were performed using SAS statistical
software, version 9.4 (SAS Institute, Cary, NC).
Patient involvement
No patients were involved in setting the research
question or the outcome measures, nor were
they involved in the design or implementation
of the study. No patients were asked to advise on
interpretation or writing up of results. There are no
plans to disseminate the results of the research to
study participants.
Results
10 year cumulative incidences of post-pregnancy
hypertension
We identified 482972 women with no cardiac or
circulatory system disorders or hypertension registered
before their first delivery and whose first delivery
occurred in or after 1995. Of these women, 23235 (4.8%)
3
RESEARCH
had a hypertensive disorder of pregnancy in their first
pregnancy, and 16611 developed hypertension during
follow-up. Women with a normotensive first pregnancy
in their 20s, 30s, or 40s had cumulative incidences of
hypertension of 4.0%, 5.7%, and 11.3%, respectively, in
the decade after delivery (fig 1, supplementary table1).
Cumulative incidence of hypertension (%)
The corresponding incidences for women whose first
pregnancy was complicated by a hypertensive disorder of
pregnancy were 13.7%, 20.3%, and 32.4%, respectively.
A similar pattern was observed in the decade after a
second pregnancy in women with two pregnancies,
where either pregnancy could have been complicated by
hypertensive disorders of pregnancy (see supplementary
table 1).
Relative risks of post-pregnancy hypertension
After excluding women with cardiovascular disease
(n=47142) or pregestational hypertension (n=20249),
1025118 women had at least one eligible pregnancy
between 1978 and 2012. Table 1 presents baseline
characteristics of the women in this cohort by history of
a hypertensive disorder of pregnancy in the pregnancy
immediately preceding study entry. In this cohort,
59319 (5.8%) women had one or more pregnancies
complicated by hypertensive disorders of pregnancy
between 1978 and 2012; 183423 developed
hypertension during follow-up (1995-2012).
Hypertensive disorder of pregnancy in most recent
pregnancyin the year after a womans latest delivery,
rates of post-pregnancy hypertension were 12-fold to
25-fold higher in women with a hypertensive disorder
of pregnancy than in women with a normotensive
pregnancy (fig 2, supplementary table 2). One to
five years post partum, rates were fourfold to 10-
fold higher in women with a hypertensive disorder
of pregnancy in the latest pregnancy. Hazard ratios
were lower thereafter, but even more than 20 years
post partum, women with a hypertensive disorder of
pregnancy in their most recent pregnancy had twice
the rate of hypertension as women whose most recent
pregnancy was normotensive. With the exception of
the first year post partum, hazard ratios for gestational
hypertension were statistically significantly higher
than hazard ratios for pre-eclampsia, regardless of
the time that had elapsed since pregnancy (fig 2,
supplementary table 2). Observed association
magnitudes did not differ statistically significantly
by severity of pre-eclampsia. Additional adjustment
for diabetes, smoking, and body mass index did not
meaningfully change the results (see supplementary
tables 3-5). In the first years after delivery, there was
a tendency towards higher rates of hypertension in
women with early onset hypertensive disorders of
pregnancy, compared with women with intermediate
onset or term hypertensive disorders of pregnancy,
but overall, hazard ratios did not vary statistically
significantly by timing of onset of the hypertensive
disorder of pregnancy (see supplementary figure 1 and
supplementary table 6).
History of hypertensive disorders of pregnancy
rates of post-pregnancy hypertension were twofold to
4
Hypertensive No hypertensive
disorder disorder
of pregnancy of pregnancy
40-49 years
30-39 years
20-29 years
95% CI
40
30
20
10
0
0 1 2 3 4 5 6 7 8 9 10
Time since first birth (years)
Fig1 | Ten year cumulative incidences of hypertension by
years since first pregnancy in women with and without
a hypertensive disorder of pregnancy, by age at first
delivery, Denmark, 1995-2012. Follow-up began in 1995
or three months post partum, whichever came later
sixfold higher in women with a history of hypertensive
disorders of pregnancy than in women the same age
with only normotensive pregnancies, with the strongest
associations in younger women (fig 3, supplementary
table 7). For all but women aged less than 30 years, a
history of gestational hypertension was more strongly
associated with post-pregnancy hypertension than a
history of pre-eclampsia, regardless of the severity of
the pre-eclampsia.
Sequence of hypertensive disorders of pregnancy
for women with at least two pregnancies, having a
hypertensive disorder of pregnancy in the first pregnancy
but not in the second was associated with a doubling
of the rate of post-pregnancy hypertension, compared
with having two normotensive pregnancies, regardless
of the time elapsed since the second pregnancy (fig 4,
supplementary table 8). The associations were even
stronger in women with a normotensive first pregnancy
and a hypertensive disorder of pregnancy in the second
pregnancy, and greatest (hazard ratio range 2.6-7.8) in
women with hypertensive disorders of pregnancy in
both pregnancies.
Sensitivity analyses
Table 2 shows the distribution of antihypertensive drug
use in the first year post partum among women with a first
delivery in or after 1995. Some hypertension identified
during this year might reflect undetected pregestational
hypertension. Consequently, we examined the effect of
excluding women who developed hypertension 3-12
months post partum on our cumulative incidences
and hazard ratios. Follow-up began one year post
partum, and in the analyses we included only women
with no hypertension 3-12 months after delivery.
Beginning follow-up one year post partum in women
free of hypertension at that time slightly reduced the
cumulative incidences for women with hypertensive
doi: 10.1136/bmj.j3078|BMJ 2017;358:j3078 | thebmj
 RESEARCH

Table 1 | Baseline characteristics of the study cohort used to estimate relative risks of post-pregnancy hypertension, at study entry, Denmark,
1995-2012*. Values are numbers (percentages) unless stated otherwise
Hypertensive disorder of pregnancy in most recent pregnancy
Characteristics Normotensive Gestational hypertension Moderate pre-eclampsia Severe pre-eclampsia
Age (years):
<20 14138 (1.4) 56 (0.6) 314 (1.3) 134 (1.8)
20-24 108333 (11.0) 810 (9.3) 3006 (12.5) 1051 (14.0)
25-29 279569 (28.4) 2554 (29.4) 7590 (31.6) 2567 (34.2)
30-34 258279 (26.2) 2277 (26.2) 6089 (25.3) 2131 (28.4)
35-39 166099 (16.9) 1485 (17.1) 3555 (14.8) 1016 (13.5)
40-44 102678 (10.4) 888 (10.2) 2155 (9.0) 427 (5.7)
45-49 44445 (4.5) 467 (5.4) 1003 (4.2) 141 (1.9)
50 11324 (1.2) 156 (1.8) 345 (1.4) 36 (0.5)
Total 984865 8693 24057 7503
Parity:
1 698753 (70.9) 7087 (81.5) 19955 (82.9) 6700 (89.3)
2 220210 (22.4) 1248 (14.4) 3281 (13.6) 634 (8.4)
3 65902 (6.7) 358 (4.1) 821 (3.4) 169 (2.3)
Total 984865 8693 24057 7503
Diabetes (type 1 or 2):
Yes 4667 (0.5) 111 (1.3) 385 (1.6) 129 (1.7)
No 980198 (99.5) 8582 (98.7) 23672 (98.4) 7374 (98.3)
Total 984865 8693 24057 7503
First trimester smoking:
Yes 162644 (23.8) 1029 (15.5) 3314 (18.2) 847 (13.3)
No 520173 (76.2) 5592 (84.5) 14879 (81.8) 5544 (86.8)
Total 682817 6621 18193 6391
Prepregnancy body mass index:
<18.5 14134 (4.5) 69 (1.8) 153 (2.1) 105 (3.5)
18.5-24 203234 (64.8) 1757 (46.4) 3608 (48.4) 1716 (57.0)
25-29 63288 (20.2) 1005 (26.5) 1983 (26.6) 687 (22.8)
30-34 22345 (7.1) 551 (14.6) 1030 (13.8) 310 (10.3)
35 10649 (3.4) 405 (10.7) 686 (9.2) 194 (6.4)
Total 313650 3787 7460 3012
*Based on full cohort of women with one or more pregnancies between 1978 and 2012, unless indicated otherwise (see and ).
History of hypertensive disorder of pregnancy in most recent pregnancy before start of follow-up. Not all women with a pregnancy complicated by a hypertensive disorder of pregnancy
had the disorder in the pregnancy immediately preceding the start of follow-up; history of a hypertensive disorder of pregnancy in the most recent pregnancy was a time dependent variable
that was reset with every pregnancy during follow-up. Therefore, although 40253 women had a history of a hypertensive disorder of pregnancy in the most recent pregnancy at the start of
follow-up (as shown in this table), 59319 had a pregnancy affected by a hypertensive disorder of pregnancy between 1978 and 2012 (as noted in the Results).
First trimester smoking status in womans first pregnancy resulting in live birth or stillbirth in 1991-2012.
Prepregnancy body mass index in womans first pregnancy resulting in live birth or stillbirth in 2004-12.

disorders of pregnancy (see supplementary table 1).
In the analyses of cumulative history of hypertensive
disorders of pregnancy, hazard ratio magnitudes
in women aged less than 35 years decreased (see
supplementary table 9), whereas hazard ratios for older
women changed little, if at all. Excluding women with
early post-pregnancy hypertension from the analyses of
hypertensive disorder of pregnancy sequence reduced
the magnitudes of hazard ratios estimated for elapsed
times of five years or less since the second pregnancy
(see supplementary table 10).
Discussion
In a large nationwide cohort, 14-32% of women with
a hypertensive disorder of pregnancy in their first
pregnancy developed hypertension in the decade
after delivery, compared with 4-11% of women
with normotensive first pregnancies. Rates of post-
pregnancy hypertension in women with a hypertensive
disorder of pregnancy in the most recent pregnancy
were 12-fold to 25-fold higher in the first year post
partum and up to 10-fold higher in the decade after
delivery, than in women without a hypertensive
disorder of pregnancy. Hypertension rates among
women with a previous hypertensive disorder of
pregnancy remained doubled more than 20 years later.
Our large cohort allowed us to examine the timing
and trajectory of post-pregnancy hypertension risk
associated with hypertensive disorders of pregnancy
in short, discrete time intervals. Including all parous
women in Denmark in the cohort minimised the
possibility of selection bias, and using prospectively
collected register data eliminated the possibility of
recall bias. Validation of diagnoses for hypertensive
disorders of pregnancy in the national patient register
has shown that although their sensitivity is variable
(such that some misclassification of pregnancies
complicated by hypertensive disorders of pregnancy
as normotensive almost certainly occurred), their
specificity is excellent (>99%),19 and the potential
impact of misclassification of hypertensive disorders
of pregnancy diagnoses on the observed associations
is therefore likely negligible. Although guidelines for
the diagnosis of hypertensive disorders of pregnancy
have changed in the past 35 years, most notably
in the past decade, the blood pressure thresholds
used to define gestational hypertension and pre-

thebmj|BMJ 2017;358:j3078 | doi: 10.1136/bmj.j3078 5

RESEARCH

Gestational hypertension Gestational hypertension

Severe pre-eclampsia Severe pre-eclampsia
Moderate pre-eclampsia Moderate pre-eclampsia
32 8

Hazard ratio (95% CI)

Hazard ratio (95% CI)

16
4
8

4
2
2

1 1
<30 30-34 35-39 40-44 45-49 50
<1

1
2
3
4
5
6
7
8

10 9
4

0
-1

-1

2
Attained age (years)

15
Years since most recent pregnancy
Fig2 | Hazard ratios for hypertension by severity of
hypertensive disorder of pregnancy (if any) in the most
recent pregnancy and time since most recent pregnancy,
among women with at least one live birth or stillbirth in
Denmark, 1978-2012. Hazard ratios compare rates of
hypertension among women with severe pre-eclampsia
(orange), moderate pre-eclampsia (green), and
gestational hypertension (black) in the latest pregnancy
with rates of hypertension in women whose most recent
pregnancy was normotensive. Follow-up began in 1995
or three months post partum, whichever came later.
Hazard ratios are adjusted for maternal age, maternal
birth year, and parity (see also supplementary figure 1
for corresponding figure with hypertensive disorders of
pregnancy classified by timing of delivery)
Fig3 | Hazard ratios for hypertension by history of
most severe hypertensive disorder of pregnancy
and attained maternal age, among women with at
least one live birth or stillbirth in Denmark, 1978-
2012. Hazard ratios compare rates of hypertension
among women with severe pre-eclampsia (orange),
moderate pre-eclampsia (green), and gestational
hypertension (black) as their most severe hypertensive
disorder of pregnancy (cumulative history over all
pregnancies) with rates of hypertension in women
whose pregnancies were all normotensive. Follow-up
began in 1995 or three months postpartum, whichever
came later. Hazard ratios are adjusted for maternal age,
maternal birth year, parity, and time since most recent
pregnancy

eclampsia, and the assumption that pre-eclampsia

typically includes proteinuria, have been fairly Hypertensive disorders of pregnancy in two successive
pregnancies
consistent. Furthermore, major changes in definition Normotensive first pregnancy followed by a hypertensive
put forward by the American College of Obstetrics and disorder of pregnancy in second pregnancy
Hypertensive disorder of pregnancy in first pregnancy
Gynecology,2 the Royal College of Obstetricians and but not in second
Gynaecologists,11 and the International Society for the 16
Hazard ratio (95% CI)

Study of Hypertension in Pregnancy,20 which allow

for conditions other than proteinuria (eg, fetal growth
restriction) to define pre-eclampsia, were published
late in the study period or after the study period ended.
Registration of filled prescriptions in the national
prescription register occurs automatically through the
personal identification number, which should ensure
that registration of all dispensed drugs in Denmark
is correct and virtually complete.17 However, women
using antihypertensive drugs for other indications
will have been misclassified as having hypertension.
Hypertension rates based on drug use also
underestimate the true rates in the study population,
since not all hypertension is detected, and not all
women with a diagnosis are treated.21 However, since
clinical awareness of the link between hypertensive
disorders of pregnancy and later cardiovascular disease
remains incomplete,9 10 22 the probability of having
a diagnosis of hypertension was unlikely to differ for
women with and without a history of hypertensive
disorders of pregnancy. Hypertension diagnosed
within one year post partum was undoubtedly also a
mixture of undetected prepregnancy hypertension and
new onset hypertension. From a clinical perspective,
however, whether hypertension detected post partum
8
4
2
1
<5 5-9 10-14 15-19 20
Years since second pregnancy
Fig4 | Hazard ratios for hypertension by hypertensive
disorder of pregnancy sequence in two first pregnancies
by number of years since the second pregnancy, among
women with at least two live births or stillbirths in
Denmark, 1978-2012. Hazard ratios compare rates
of hypertension among women with a hypertensive
disorder of pregnancy in first pregnancy but not in
the second (green), a normotensive first pregnancy
followed by a hypertensive disorder of pregnancy in
second pregnancy (red), and hypertensive disorders
of pregnancy in two successive pregnancies (purple),
with rates in women with two successive normotensive
pregnancies. Follow-up began in 1995 or 3 months after
the second delivery, whichever came later. Hazard ratios
are adjusted for maternal age, maternal birth year, and
parity

6 doi: 10.1136/bmj.j3078|BMJ 2017;358:j3078 | thebmj


Table2 | Antihypertensive drug use in the year after first delivery in women without previous antihypertensive drug use, by history of hypertensive
disorders of pregnancy in Denmark, 1995-2012*
Antihypertensive drug use in year after delivery
Hypertensive disorder of <3 months post
pregnancy in first pregnancy Total No of women None partum
Yes 23826 21216 1915
No 461290 458995 1335
*485116 women had their first delivery in or after 1995; these women are included in the table, which also includes information on use of antihypertensive drugs from delivery until one year
post partum. Of these women, 2144 did not enter follow-up (which began three months after delivery) because they were censored before this point due to, for example, development of heart
disease. Consequently, in the text we state that the estimates of cumulative incidence are based on a slight smaller number of women (n=482972).
predated the pregnancy or is a sequela of hypertensive
disorders of pregnancy, is of minor importance; either
way, women with a previous hypertensive disorder of
pregnancy have much higher rates of hypertension
than their peers, and clinical contact with these
women therefore represents an important opportunity
for detection. Adjustment for diabetes and smoking
did not change the observed associations, suggesting
that these variables could not explain our findings.
Although adjustment for body mass index also did
not substantially change our results, these analyses
were underpowered and we cannot exclude the
possibility that body mass index might play a role
in the association between hypertensive disorders
of pregnancy and post-pregnancy hypertension. We
also cannot discount the possibility that unmeasured
risk factors for hypertension (eg, alcohol and salt
intake, stress) might have confounded the observed
associations, but this is unlikely, since to do so
these factors would also have to be associated with
hypertensive disorders of pregnancy.
Most studies assessing hypertension risk after a
hypertensive disorder of pregnancy have reported
risks averaged over follow-up periods ranging from a
few years to decades (eg,3-5 13). Apart from one study
examining hazard ratios for hypertension in five year
intervals after delivery,12 studies have not focused
on hypertension risk specifically by time since a
pregnancy complicated by a hypertensive disorder
of pregnancy that is, on the timing of hypertension
onset and how the risk of hypertension changes
over time. Our work indicates that the immediate
postpartum years are important: the increased risks of
hypertension associated with hypertensive disorders
of pregnancy already exist shortly after an affected
pregnancy. Furthermore, for women aged more than
30 years, gestational hypertension was more strongly
associated with post-pregnancy hypertension than
was pre-eclampsia, which is consistent with findings
from some,23 24 but not all,4 13 25 previous studies and
is thought to reflect aetiological differences between
gestational hypertension and pre-eclampsia.2627
Hypertension during and after pre-eclampsia might
be a marker of wider multiorgan system disruption,
whereas the mechanisms underlying gestational
hypertension may be more narrowly focused. An
excess of undetected prepregnancy hypertension
in women with gestational hypertension could also
explain higher rates of hypertension immediately after
an affected pregnancy, but differential detection of
thebmj|BMJ 2017;358:j3078 | doi: 10.1136/bmj.j3078
RESEARCH
<3 months and 3-12 months post
partum 3-12 months post partum only
513 182
276 684
prepregnancy hypertension depending on later type of
hypertensive disorder of pregnancy seems implausible.
Furthermore, such an excess would not explain why
risks remained larger for gestational hypertension than
for pre-eclampsia decades after delivery.
Women with hypertensive disorders of pregnancy
had much higher 10 year risks of post-pregnancy
hypertension than did women of the same age with
normotensive pregnancies. The substantial absolute
risks of hypertension in the decade after a hypertensive
disorder of pregnancy indicate that the processes
linking hypertensive disorders of pregnancy and
hypertension are already operational during or shortly
after pregnancy. Whether the risk of hypertension
(and later cardiovascular disease) in women with
hypertensive disorders of pregnancy is due to common
predisposing factors or to a downstream effect of
pathophysiological processes specific to hypertensive
disorders of pregnancy, is currently the focus of
debate.28-30 Either hypothesis could explain the finding
that women with hypertensive disorders of pregnancy
in two pregnancies had higher risks of hypertension
than women with one affected pregnancy. However,
the finding that, among women with a hypertensive
disorder of pregnancy in only one of two pregnancies,
a hypertensive disorder of pregnancy in the second
pregnancy was more strongly associated with later
hypertension than a hypertensive disorder of pregnancy
in the first pregnancy, argues against the importance
of processes initiated by hypertensive disorders of
pregnancy. If post-partum hypertension was the direct
consequence of the hypertensive disorder of pregnancy
itself, one would expect the effect of a single affected
pregnancy to be the same regardless of sequence, after
taking into account time since most recent pregnancy.
The study results suggest that the as yet poorly
understood processes leading to hypertension related
to hypertensive disorders of pregnancy begin while
the affected women are still relatively young. At
the very least, initiation of regular blood pressure
assessments soon after a pregnancy complicated
by a hypertensive disorder of pregnancy is essential
for prompt identification of hypertension in these
women. Because the number of women potentially at
risk of hypertension related to hypertensive disorders
of pregnancy is large, and routine follow-up could
conceivably last years or even decades, an algorithm
to identify those at greatest risk (the subgroup most
likely to benefit from screening) is urgently needed;
identification of biomarkers that predict which women
7
RESEARCH
will develop hypertension after hypertensive disorders
of pregnancy would be even more useful. Furthermore,
the degree to which early identification and timely
treatment of hypertension in women with a history of
hypertensive disorders of pregnancy can reduce their
burden of cardiovascular disease, is also unknown
and needs to be clarified. In the quest to minimise the
lifetime impact of hypertensive disorders of pregnancy
on womens post-pregnancy health, we now need data,
particularly from randomised clinical trials, to support
clinical decision making and policy on clinical follow-
up for women with a history of hypertensive disorders
of pregnancy.
Contributors: IB conceived the study, contributed to the study
design, classified register data, interpreted study results, and drafted
the paper. SB designed the study; planned the statistical analyses;
obtained, linked, and analysed the data; interpreted the study
results; and revised the paper. MM, JAL, and HB contributed to the
study design, interpreted the study results, and revised the paper.
JW designed the study, planned the statistical analyses and oversaw
their conduct, interpreted the study results, and revised the paper.
BT conceived the study, interpreted the study results, and revised
the paper. HAB designed the study, planned the statistical analyses,
obtained the data, interpreted the study results, and drafted and
revised the paper. She is guarantor of the paper. All authors had full
access to all of the data in the study and can take responsibility for the
integrity of the data and the accuracy of the data analysis.
Funding: This study was funded by the Danish Council for
Independent Research and the Danish Heart Association. Neither
had a role in the design and conduct of the study; collection,
management, analysis, and interpretation of the data; preparation,
review, or approval of the manuscript; or the decision to submit the
manuscript for publication. The researchers acted independently from
the study sponsors in all aspects of this study.
Competing interests: All authors have completed the ICMJE uniform
disclosure form at www.icmje.org/coi_disclosure.pdf and declare:
IB and SB were supported by a grant from the Danish Council for
Independent Research; IB also received grant support from the
Danish Heart Association; no other support for the submitted work;
no financial relationships with any organisations that might have an
interest in the submitted work in the previous three years; no other
relationships or activities that could appear to have influenced the
submitted work.
Ethical approval: Studies based solely on data from the Danish
national registers do not require approval from the Danish research
bioethics committees, as study participants are never contacted, and
consent is not required for the use of register information. The studys
use of register data was covered by the approval extended by the
Danish Data Protection Agency to all register based studies conducted
by Statens Serum Institut (approval No 2015-57-0102).
Data sharing: This study is based on Danish national register data.
These data do not belong to the authors but to the Danish Ministry
of Health, and the authors are not permitted to share them, except
in aggregate (as, for example, in a publication). However, interested
parties can obtain the data on which the study was based by
submitting a research protocol to the Danish Data Protection Agency
(Datatilsynet) and then, once Data Protection Agency permission has
been received, applying to the Ministry of Healths Research Service
(Forskerservice) at forskerservice@ssi.dk.
Transparency: The guarantor (HAB) affirms that the manuscript is
an honest, accurate, and transparent account of the study being
reported; that no important aspects of the study have been omitted;
and that any discrepancies from the study as planned have been
explained.
This is an Open Access article distributed in accordance with the
Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license,
which permits others to distribute, remix, adapt, build upon this work
non-commercially, and license their derivative works on different
terms, provided the original work is properly cited and the use is non-
commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
1 Abalos E, Cuesta C, Grosso AL, Chou D, Say L. Global and regional
estimates of preeclampsia and eclampsia: a systematic review.
Eur J Obstet Gynecol Reprod Biol 2013;170:1-7. doi:10.1016/j.
ejogrb.2013.05.005
8
2 American College of Obstetricians and Gynecologists. Task Force on
Hypertension in Pregnancy. Hypertension in pregnancy. Report of the
American College of Obstetricians and Gynecologists Task Force on
Hypertension in Pregnancy. Obstet Gynecol 2013;122:1122-31.
3 Bellamy L, Casas J-P, Hingorani AD, Williams DJ. Pre-eclampsia and
risk of cardiovascular disease and cancer in later life: systematic
review and meta-analysis. BMJ 2007;335:974. doi:10.1136/
bmj.39335.385301.BE
4 Lykke JA, Langhoff-Roos J, Sibai BM, Funai EF, Triche EW, Paidas MJ.
Hypertensive pregnancy disorders and subsequent cardiovascular
morbidity and type 2 diabetes mellitus in the mother. Hypertension
2009;53:944-51. doi:10.1161/HYPERTENSIONAHA.109.130765
5 Brown MC, Best KE, Pearce MS, Waugh J, Robson SC, Bell R.
Cardiovascular disease risk in women with pre-eclampsia:
systematic review and meta-analysis. Eur J Epidemiol 2013;28:1-19.
doi:10.1007/s10654-013-9762-6
6 Mosca L, Benjamin EJ, Berra K. American Heart Association.
Effectiveness-based guidelines for the prevention of cardiovascular
disease in women--2011 update: a guideline from the American
Heart Association. J Am Coll Cardiol 2011;57:1404-23.
doi:10.1016/j.jacc.2011.02.005
7 Piepoli MF, Hoes AW, Agewall S. Authors/Task Force Members.
2016 European guidelines on cardiovascular disease prevention in
clinical practice: The Sixth Joint Task Force of the European Society of
Cardiology and other societies on cardiovascular disease prevention
in clinical practice. Eur Heart J 2016;37:2315-81. doi:10.1093/
eurheartj/ehw106
8 Wilkins-Haug L, Celi A, Thomas A, Frolkis J, Seely EW. Recognition
by womens health care providers of long-term cardiovascular
disease risk after preeclampsia. Obstet Gynecol 2015;125:1287-92.
doi:10.1097/AOG.0000000000000856
9 Young B, Hacker MR, Rana S. Physicians knowledge of future
vascular disease in women with preeclampsia. Hypertens Pregnancy
2012;31:50-8. doi:10.3109/10641955.2010.544955
10 Heidrich M-B, Wenzel D, von Kaisenberg CS, Schippert C, von
Versen-Hynck FM. Preeclampsia and long-term risk of cardiovascular
disease: what do obstetrician-gynecologists know? BMC Pregnancy
Childbirth 2013;13:61. doi:10.1186/1471-2393-13-61
11 National Collaborating Centre for Womens and Childrens Health.
Hypertension in pregnancy: the management of hypertensive
disorders during pregnancy. RCOG Press, 2011.
12 Engeland A, Bjrge T, Klungsyr K, Skjrven R, Skurtveit S, Furu K.
Preeclampsia in pregnancy and later use of antihypertensive drugs.
Eur J Epidemiol 2015;30:501-8. doi:10.1007/s10654-015-0018-5
13 Veerbeek JHW, Hermes W, Breimer AY. Cardiovascular disease risk
factors after early-onset preeclampsia, late-onset preeclampsia, and
pregnancy-induced hypertension. Hypertension 2015;65:600-6.
doi:10.1161/HYPERTENSIONAHA.114.04850
14 Schmidt M, Pedersen L, Srensen HT. The Danish Civil Registration
System as a tool in epidemiology. Eur J Epidemiol 2014;29:541-9.
doi:10.1007/s10654-014-9930-3
15 Lynge E, Sandegaard JL, Rebolj M. The Danish National
Patient Register. Scand J Public Health 2011;39(Suppl):30-3.
doi:10.1177/1403494811401482
16 Knudsen LB, Olsen J. The Danish Medical Birth Registry. Dan Med Bull
1998;45:320-3.
17 Kildemoes HW, Srensen HT, Hallas J. The Danish National
Prescription Registry. Scand J Public Health 2011;39(Suppl):38-41.
doi:10.1177/1403494810394717
18 Lin DY, Wei LJ, Ying Z. Checking the Cox model with cumulative
sums of martingale-based residuals. Biometrika 1993;80:557-72
doi:10.1093/biomet/80.3.557.
19 Klemmensen AK, Olsen SF, Osterdal ML, Tabor A. Validity of
preeclampsia-related diagnoses recorded in a national hospital
registry and in a postpartum interview of the women. Am J Epidemiol
2007;166:117-24. doi:10.1093/aje/kwm139
20 Tranquilli AL, Dekker G, Magee L. The classification, diagnosis and
management of the hypertensive disorders of pregnancy: A revised
statement from the ISSHP. Pregnancy Hypertens 2014;4:97-104.
21 Joffres M, Falaschetti E, Gillespie C. Hypertension prevalence,
awareness, treatment and control in national surveys from England,
the USA and Canada, and correlation with stroke and ischaemic
heart disease mortality: a cross-sectional study. BMJ Open
2013;3:e003423. doi:10.1136/bmjopen-2013-003423
22 MacDonald SE, Walker M, Ramshaw H, Godwin M, Chen XK, Smith
G. Hypertensive disorders of pregnancy and long-term risk of
hypertension: what do Ontario prenatal care providers know, and
what do they communicate? J Obstet Gynaecol Can 2007;29:705-10.
doi:10.1016/S1701-2163(16)32601-9
23 Magnussen EB, Vatten LJ, Smith GD, Romundstad PR. Hypertensive
disorders in pregnancy and subsequently measured cardiovascular
risk factors. Obstet Gynecol 2009;114:961-70. doi:10.1097/
AOG.0b013e3181bb0dfc
24 Nisell H, Lintu H, Lunell NO, Mllerstrm G, Pettersson E. Blood
pressure and renal function seven years after pregnancy complicated
doi: 10.1136/bmj.j3078|BMJ 2017;358:j3078 | thebmj

by hypertension. Br J Obstet Gynaecol 1995;102:876-81.
doi:10.1111/j.1471-0528.1995.tb10874.x
25 Wilson BJ, Watson MS, Prescott GJ. Hypertensive diseases of
pregnancy and risk of hypertension and stroke in later life:
results from cohort study. BMJ 2003;326:845. doi:10.1136/
bmj.326.7394.845
26 Egeland GM, Klungsyr K, yen N, Tell GS, Nss , Skjrven R.
Preconception cardiovascular risk factor differences between
gestational hypertension and preeclampsia: Cohort Norway
Study. Hypertension 2016;67:1173-80. doi:10.1161/
HYPERTENSIONAHA.116.07099
27 Noori M, Donald AE, Angelakopoulou A, Hingorani AD, Williams
DJ. Prospective study of placental angiogenic factors and maternal
vascular function before and after preeclampsia and gestational
hypertension. Circulation 2010;122:478-87. doi:10.1161/
CIRCULATIONAHA.109.895458
No commercial reuse: See rights and reprints http://www.bmj.com/permissions
RESEARCH
28 Romundstad PR, Magnussen EB, Smith GD, Vatten LJ.
Hypertension in pregnancy and later cardiovascular risk: common
antecedents? Circulation 2010;122:579-84. doi:10.1161/
CIRCULATIONAHA.110.943407
29 Rich-Edwards JW, Fraser A, Lawlor DA, Catov JM. Pregnancy
characteristics and womens future cardiovascular health: an
underused opportunity to improve womens health? Epidemiol Rev
2014;36:57-70. doi:10.1093/epirev/mxt006
30 Staff AC, Redman CWG, Williams D. Global Pregnancy Collaboration
(CoLab). Pregnancy and long-term maternal cardiovascular health:
progress through harmonization of research cohorts and biobanks.
Hypertension 2016;67:251-60.
Supplementary material:
Supplementary tables 1-10 and supplementary figure
Subscribe: http://www.bmj.com/subscribe